DESCRIPTION: Binge Eating Disorder (BED), a new DSM-IV eating disorder diagnostic category, is a serious, prevalent, and refractory public health problem. Current treatment approaches for BED have utilized interventions found effective for bulimia nervosa, another eating disorder characterized by binge eating. Initial controlled trials with BED have found that cognitive behavioral therapy (CBT) and certain antidepressants may produce short-term improvements in binge eating. The proposed study represents a controlled clinical trial for BED to test the efficacy of fluoxetine treatment and CBT and allow for testing of the relative efficacy of these treatments either alone or in combination. Subjects will be 120 patients with DSM-IV BED who are randomly assigned to one of four treatment conditions for 16 weeks: (1) fluoxetine treatment (60 mg/day), (2) pill placebo, (3) fluoxetine treatment (60 mg/day) combined with CBT, and (4) pill placebo combined with CBT. The fluoxetine and pill placebo treatments will be administered in double blind fashion. Repeated assessments will occur at 4, 8, 12, and 16 weeks (end of treatment), and posttreatment (6 and 12 months). The primary aim is to test the efficacy of fluoxetine treatment and CBT and test the relative efficacy of these treatments either alone or in combination. The secondary goals are to evaluate (a) the maintenance of clinical effects during the one-year posttreatment period, (b) differences in the patterns and types of clinical changes during the treatments, (c) whether differential short-term symptom changes are associated with long-term maintenance, and (d) whether pre-treatment patient variables are predictors of outcome. In addition to informing the treatment of BED, the study will provide preliminary data concerning how different treatments work.